1. Technical Field
The present disclosure relates to addressing of server modules and more specifically to reducing the number of pins required for addressing of server modules.
2. Introduction
Modular server designs can provide improved flexibility with regards to function, operation, power management, and repair. However, these improvements come at a cost of having to connect, communicate, and at times restore communications on a module-by-module basis. For example, every time a new blade is inserted into a blade server, the server must identify the address which will be used for that blade, communicate the address to the blade, and ensure the blade will effectively communicate with the server. In other configurations (such as configurations where the blade has particular configurations or functions), the address of the blade can be specific to the blade, such that if the blade were to move locations within the server (or be moved to another server), the address for the blade remains the same.
To establish communications between the removable modules and the server, servers and modules typically have a number of pins used exclusively for identification purposes, where the number of pins corresponds to the maximum number of modules which can be housed in the server. More specifically, the number of pins required for module identification generally operates on a “base-two” requirement, such that if a server can hold eight modules (23), three pins are required, whereas if a server can hold up to 32 modules (25), five pins are required. Because of this, the number of input/output pins of the module which are dedicated to module identification and addressing must increase as the capacity of the server increases. Increasing the number of pins dedicated to module identification in turn reduces total bandwidth and capacity for other communications between the server and the server module.